oakla, ’ash’ and ’elm’. Among the softwoods, in the twenty-year period ’spruce’ made the greatest overall contribution (35.4%) (it was probably mostly black spruce, though red and white spruce may also have been used), followed by tamarack (always called ’juniper' on the forms) (19.4%), and ’pine’ (sometimes specified as red or white pine) (4.5%), with even smaller frequencies for ’cedar’, ’hemlock’ and ’pitch pine’ — this last could only have been imported. However, as is evident in Table 4-1, these overall values mask important changes in the frequencies of the usage of these woods in the twenty years between 1856 and 1876: in 1856 hardwoods predominated, with an overall frequency of 56.8%, but by 1876 they had fallen to just 21.7%. Concurrently, the frequency of the usage of softwoods rose from 43.2% in 1856 to 78.3% in 1876, this rise being almost entirely accounted for by ’spruce’, which rose threefold from a frequency of 23.9% to 75.1%, while the use of all other woods (both hard and soft) showed a marked decline over the period. The pattern in the use of tamarack is of particular interest: Table 4-1 shows that its use in island ships was greatest in the 1861 sample, thereafter declining to insignificance by 1876. As noted in the main reportg, from about the 18405 tamarack was the preferred wood for ship-building, being used, as Table 4-4 shows, for all parts of the vessel except the keel, the rider keelson and the decks. Having completed this analysis using sixty vessels, it was decided to extend the sample to every vessel inspected in each of the same five years. Based on an impression of the predominant wood- types listed on the report, my aim was to place each vessel in one of five categories: ’juniper’ vessels, ‘spruce’ vessels, vessels of ’mixed woods’ (hardwoods making a big contribution to these). and vessels of mixed woods in which there was a large amount of spruce.10 The results of this 8 As is evident from Table 4—4, ‘oak' was used especially for the rudder and the Windlass, two hard-wearing components that were not built into the structure of the ship. The wood for both, or even the completed article, could easily have been imported, either as white oak from elsewhere in North America (Quebec is mentioned as a source in some newspaper accounts), or European oak from England. 9 See footnote 691, page 103 1° Since the aim was to carry out a quick survey of all the vessels of those years (some 257 vessels), quantitative counts of woody components were not carried out, which means there was an element of subjectivity in the placing of some of the vessels into the 276 extended survey, which are presented in Table 4-2, are similar to those of the more quantitative survey just discussed. They indicate that, whereas in 1856 all vessels were either of ’juniper’ or of ’mixed woods’, by 1876 all but five of the 41 vessels inspected in that year were either ’spruce’ vessels or of mixed woods with a high component of spruce. Thus my more objective and quantitative study of the records supports the casual observation of Greenhill and Giffard, based on a few of the ships of James Yeo, that in the twenty year period there was a decline in the use of hardwoods and tamarack, and an increase in the use of spruce, and that this change occurred over the whole island. This shift in the usage by island ship-builders of the various types of wood requires an explanation. Greenhill and Giffard attributed the changes to the ’exhaustion’ of these more preferred woods, and undoubtedly this is the explanation for tamarack, ships of which throughout the period were always given a higher rating (by about two years) by the Lloyd’s inspector than ships built of any other wood. We may thus assume that, had it been available in the 18705, tamarack would have been used by ship—builders, and therefore the decline in its use from its peak in the 1861 sample, when almost 60% of the ships were ’juniper’ ships, to just one ship fifteen years later can only have been due to a decline in the availability of the wood. This, as Greenhill and Giffard suggest, may also be the explanation for the decline in the use of the hardwoods, though the Lloyd’s reports reveal a complicating factor: in the 18705, vessels constructed of spruce were rated higher (usually at 7A) by the Lloyd’s inspector than were those of mixed woods (usually 5A), indicating that spruce vessels were viewed as having a greater longevity, and this could have led to ship-builders preferring five classes. ‘Juniper’ vessels were easily recognized, since tamarack was listed for all of the components except the keel and the deck (in Table 4—3 there are 17 such ships, each with 21 or more structural components for which tamarack is listed — the fact that all 'juniper‘ vessels were rated by the inspector at least two classes higher than any other vessel-type, also provided a useful marker for the class.) Those vessels classed as ‘mixed’ had a high hardwood component throughout (usually birch and beech), though there was also the customary use of softwoods (especially spruce and tamarack) in the deck, as well as some other parts of the vessels. ‘Spruce' vessels had almost all of their components of spruce except for those parts always in contact with water (i.e. the floors. the first futtocks. the keel and deadwood, as well as the lowest levels of the exterior planking). This left some vessels that had a high component of spruce but had a slightly higher frequency of hardwoods than those classified as “spruce vessels" — these were placed in a class called “vessels of mixed woods, with a large amount of spruce".